Classifications

• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
  • G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
  • G10L21/0208—Noise filtering
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
  • H04B1/06—Receivers
  • H04B1/10—Means associated with receiver for limiting or suppressing noise or interference
  • H04B1/1027—Means associated with receiver for limiting or suppressing noise or interference assessing signal quality or detecting noise/interference for the received signal

Definitions

• the present invention relates to a noise suppression device that suppresses a noise component from a signal mixed with noise, and particularly to a noise suppression device in a receiver such as radio wave communication.
• FIG. 1 is a diagram illustrating a configuration of a noise suppression device using a conventional spectral subtraction method.
• FIG. 2 is a diagram showing a configuration of a noise suppression device using a conventional Wiener filter.
• the spectral subtraction method or Wiener filter estimates (measures) the noise spectrum in the noise interval, which is the time interval, without speech. Then, noise suppression is performed based on the basic principle of subtracting the input signal power from the noise.
• any suppression method in order to adjust the amount of noise suppression, for example, it is common to multiply the noise spectrum estimated according to the following equation by a suppression coefficient a that is a magnification variable.
• X (f) is the input signal spectrum
• N (f) is the noise spectrum
• Y (f) is the signal spectrum after noise suppression
• I I is the absolute value of the signal.
• f represents the frequency.
• a musical noise peculiar to the spectral subtraction method is generated. Although there is an improvement in the SZN ratio, it is a disadvantage that musical noise such as “shurushuru” remains after suppressing the noise.
• the cause of the degradation of the noise spectrum estimation accuracy is caused by non-uniform environmental noise. Since the noise spectrum is generally not constant, it can be expected that a higher noise suppression effect can be obtained if the suppression coefficient ⁇ is variable according to the environment. For example, the current input signal A noise suppression amount is calculated based on the standard deviation of noise that is obtained only from the spectrum and the average spectrum of noise, and has been proposed that further corresponds to actual environmental noise (see, for example, Patent Document 1). However, even the method of Patent Document 1 cannot cope with transient noise that is uncorrelated with the past noise spectrum.
• Patent Document 2 JP 2003-316381
• Patent Document 2 Japanese Patent No. 2760240
• the suppression process is inevitably a process with a certain time width.
• the signal band is limited to 15 kHz. Therefore, assuming that sampling is performed at 32 kHz, if spectrum conversion is performed at 256 points, suppression processing is performed in units of 8 ms.
• the received electric field strength changes from moment to moment.
• the pulse noise generated by the dip of the electric field strength in the multipath environment is about several ⁇ s.
• the noise in the 8ms frame is uniformly suppressed to suppress the noise of the number / zs in the 8ms frame, the signal other than the signal to which the pulse noise is added may be excessively removed, causing the generation of musical noise. End up.
• the present invention solves the above-described problem, and can effectively suppress stationary noise due to a decrease in electric field strength, and can also effectively suppress transient environmental noise.
• An object of the present invention is to provide a noise suppression device.
• the noise suppression device of the present invention stores the signal strength value of the digital input signal according to the signal strength of the input signal and the storage means for storing the signal strength value.
• Frame forming means for forming a frame which is a collection of signal values including the signal value, and noise mixed in the input signal in the frame according to the signal strength of the formed frame
• Noise estimation means for estimating the noise pattern
• noise suppression means for suppressing the noise included in the formed frame using the estimated noise pattern.
• the present invention may be configured as an FM and AM receiving device having such a noise suppression device that can only be realized as such a noise suppression device, and such noise suppression is also possible. It can be realized as a noise suppression method using characteristic means of the apparatus as steps, or as a program for causing a computer to execute these steps. Needless to say, such a program can be distributed via a recording medium such as a CD-ROM or a transmission medium such as the Internet.
• non-stationary noise By forming a frame that matches the change in the received electric field strength, even non-stationary noise can be regarded as stationary noise only within a shortened frame. The error can be reduced. At the same time, it is estimated whether the noise is stationary or non-stationary by monitoring changes in the received electric field strength. In the case of non-stationary noise, the noise suppression effect is improved by performing stronger suppression processing within a short frame. Can be improved.
• FIG. 1 is a diagram showing the configuration of a conventional noise suppression apparatus using a spectral subtraction method.
• FIG. 2 is a diagram showing a configuration of a noise suppression device using a conventional Wiener filter.
• FIG. 3 is a block diagram showing a configuration of a noise suppression device in the present embodiment.
• FIG. 4 is a block diagram showing an example of the configuration of a radio equipped with the noise suppression device shown in FIG.
• FIG. 5 is a flowchart showing an example of a frame forming procedure performed by the electric field strength analysis unit shown in FIG.
• FIG. 6 is a diagram showing the relationship between the change in electric field strength and the frame separation.
• FIG. 7 is a diagram showing an example of a more detailed configuration of the electric field strength analysis unit shown in FIG.
• FIG. 8 is a diagram showing an example of a configuration in which a frame is formed with a frame length specified by the electric field strength analysis unit, and spectrum conversion is performed with the same length as the formed frame.
• FIG. 9 is a diagram showing an example of a detailed configuration of the noise estimation unit shown in FIG.
• Fig. 10 is a graph showing the noise spectrum of the FM receiver when the received electric field strength is OdB with the FFT frame length changed.
• FIG. 11 is a block diagram showing an example of a detailed configuration in the case of suppressing noise of an input signal by subtracting the noise spectrum from the input spectrum in the state of the frequency spectrum.
• FIG. 12 is a block diagram showing an example of a detailed configuration when noise suppression is performed after inversely converting a noise spectrum and an input signal into a time domain signal.
• FIG. 13 is a diagram illustrating an example of processing in the frame forming unit when a fixed-length frame is formed by zero-filling both ends of the determined frame length.
• FIG. 14 is a flowchart showing a procedure in the frame forming unit when a lower limit is set for the frame length.
• FIG. 15 is a block diagram showing an example of a configuration in the case where the field strength is compared for each sample without obtaining the average of the field strength in the averaged frame.
• Fig. 16 is a diagram showing an example in which the electric field strength of the input signal is divided into several zones corresponding to the noise pattern peculiar to the electric field strength.
• FIG. 17 is a flowchart showing an example in which the field strength of each sample of the input signal is examined and a new frame is formed each time the field strength zone changes.
• FIG. 18 is a block diagram showing an example of the configuration of a noise suppression device that estimates noise patterns mixed in an input signal using signal strength levels as a guide instead of electric field strength and suppresses noise. .
• FIG. 3 is a block diagram showing the configuration of the noise suppression apparatus according to the present embodiment.
• the noise suppression apparatus 100 according to the first embodiment pays attention to the fact that the noise pattern changes according to the degree of change in the electric field strength, and collects a set of samples in which the change in electric field strength is within a certain range. After converting the spectrum for each frame, the difference from the noise spectrum corresponding to each frame is output. Normally, a fixed number of consecutive samples, which are preliminarily determined, are used as a frame, and a power for performing processing such as frequency conversion and quantization. Here, a variable length frame is used.
• FIG. 4 is a block diagram illustrating an example of a configuration of a radio including the noise suppression device 100 illustrated in FIG.
• the radio includes an antenna 101, a tuner 102, an AZD modification 103, a noise suppression device 100, and a speaker 104.
• the antenna 101 receives an analog FM signal or AM signal transmitted from a broadcasting station.
• the tuner 102 separates and extracts, for example, an audio signal from a carrier wave having a desired frequency.
• the AZD conversion 103 converts the separated and extracted analog signal into a digital signal.
• the noise suppression device 100 suppresses the noise of the digital signal.
• the speaker 104 converts the noise-suppressed digital signal into an analog audio signal and outputs audio.
• the noise suppression device 100 shown in FIG. 3 includes a frame forming unit 1, a spectrum conversion unit 2, an electric field strength analysis unit 3, a noise estimation unit 4, and a noise suppression unit 5.
• the frame forming unit 1 corresponds to a part of “a frame forming unit that forms a frame that is a collection of signal values including the stored signal values according to the signal strength of the input signal”. Then, from the time-series signal input as a digital value by an external AZD change, a frame having a time length necessary for processing by the subsequent spectrum conversion unit 2 is formed.
• the spectrum converter 2 converts the input time-series signal of one frame into a frequency spectrum corresponding to the frame length of the signal. Alternatively, it is converted to a power spectrum according to the subsequent processing.
• Time-series signal force Fast Fourier transform for conversion to frequency spectrum (Hereinafter referred to as “FFT”) is a generally known force.
• FFT Fast Fourier transform for conversion to frequency spectrum
• the method is not limited.
• the length of the frame formed by the frame forming unit 1 is determined by the electric field strength analyzing unit 3.
• the electric field strength analysis unit 3 includes: "a storage means for storing a signal strength value of a digital input signal, and a signal value including the signal value stored according to the signal strength of the input signal input.
• the frame length (number of samples n) at the time of spectrum conversion of the input signal X is determined according to the change in the electric field strength d. Specifically, while the electric field strength changes within a predetermined range, the time-series signals input during that time are set to the same frame, and conversely, when the electric field strength changes more than a certain amount, Let the input signal be another frame. At the same time, the electric field strength analysis unit 3 determines the spectrum length n ′ of the noise spectrum estimated by the noise estimation unit 4.
• the noise estimation unit 4 corresponds to “a noise estimation unit that estimates a noise pattern mixed in the input signal in the frame according to the signal strength of the formed frame”, and the input electric field strength. Estimate and output the noise spectrum for.
• the output noise spectrum may be a frequency spectrum or a power spectrum, depending on the subsequent processing.
• the number of points of the noise spectrum output from the noise estimation unit 4 is also determined by the electric field strength analysis unit 3 as in the case of frame formation of the input signal.
• the spectrum length of the noise spectrum is also set to a spectrum length consisting of the same number of points as the frame length of the input signal, for example.
• the noise suppression unit 5 corresponds to “a noise suppression unit that suppresses the noise included in the formed frame using an estimated noise pattern”.
• the input signal spectrum and estimated noise spectrum are used to determine the amount of suppression, and the determined amount of suppression is subtracted from the input signal, for example, to suppress the noise component contained in the input signal, and the signal after suppression Is output.
• FIG. 5 is a flowchart showing an example of a frame forming procedure performed by the electric field strength analysis unit 3 shown in FIG.
• FIG. 6 is a diagram showing the relationship between the change in electric field strength and frame separation.
• FIG. 7 is a diagram showing an example of a more detailed configuration of the electric field strength analysis unit 3 shown in FIG.
• Electric field strength analysis unit 3 detects the electric field strength of points that are input one after another And a processing unit for determining the length of each frame.
• the electric field intensity analysis unit 3 includes an electric field intensity averaging unit 31, an electric field intensity change detection unit 32, and a frame length determination unit 33 therein.
• the electric field intensity averaging unit 31 corresponds to “a signal intensity calculating unit that calculates an average value of the signal intensity in each small frame, with a small frame having a predetermined length that can form a frame as a unit”.
• the electric field strength change detecting unit 32 corresponds to “a signal strength change detecting means for detecting a change in the strength of the input signal from the strength of the input signal input”. Further, it corresponds to an “average value comparison unit that compares the average value of the signal strength of the small frame with the average value of the signal strength of the immediately preceding small frame”.
• the signal value of a point constituting the minimum frame (for example, tl to t2 in Fig. 6) is input to the electric field intensity analysis unit 3 (S301), the same number of electric fields as the number of samples of the number of points is input. Intensity d is accumulated. The electric field strength averaging unit 31 averages these accumulated electric field strengths (S302).
• the predetermined number of points is the number of points included in the minimum frame time in which the frame forming unit 1 forms a frame (hereinafter referred to as “averaged frame”).
• the number of points is a power of 2 and is preferably about 64 points on average.
• the electric field strength d of the new averaged frame (for example, t2 to t3) is averaged (S303), and the electric field strength change detection unit 32 calculates the previous average (of the previous averaged frame (tl to t2)). Is compared with the electric field strength (S304). As a result of the comparison, if the change in the averaged electric field strength is less than (or less than) a certain threshold value, the electric field strength change detection unit 32 determines the averaged electric field strength of the next averaged frame (for example, t3 to t4). The comparison is continued (S303, S304).
• the average electric field strength of the averaged frames t6 to t7 and the average value of the electric field strengths of the averaged frames t7 to t8 are compared. If the change in intensity exceeds (or exceeds) the threshold, frame formation is performed with successive averaged frames up to the averaged frame (t7) (S306).
• the electric field strength change detection unit 32 may compare the continuous averaged frames with the first averaged frame (in the immediately preceding frame), for example.
• the current averaged frame power may also be compared to the averaged frame at the beginning of several previous frames, for example, in the range V, not exceeding 1024 points! / ⁇ .
• the electric field strength of the current averaged frame is assigned to the current averaged frame. Compared with the averaged frame at the beginning of the previous frame in time, the electric field strength changes gently, and even if there is no big difference between adjacent averaged frames, There is an effect that a change in electric field strength can be detected.
• the comparison with the threshold value may be a difference between the average electric field strengths of the two averaged frames or a ratio.
• the comparison with the threshold value may be a difference between the average electric field strengths of the two averaged frames or a ratio.
• FIG. 8 is a diagram illustrating an example of a configuration in which a frame is formed with a frame length specified by the electric field strength analysis unit 3 and spectrum conversion is performed with the same length as the formed frame.
• the spectrum conversion unit 2 converts the input signal X128 input into 128 high-speed Fourier transform coefficients FTT, and outputs the absolute value I X I 128 to the noise suppression unit 5.
• FIG. 9 is a diagram illustrating an example of a detailed configuration of the noise estimation unit 4 illustrated in FIG.
• the noise estimation unit 4 is a processing unit that outputs a noise pattern at the electric field strength based on the electric field strength d of each point to which an external force is also input and the frame length ⁇ ′ input from the electric field strength analysis unit 3.
• the noise estimation unit 4 includes a noise pattern storage unit 42 therein, and stores noise patterns for various electric field strengths.
• the noise pattern is preliminarily stored and stored in the noise pattern storage unit 42, but this noise pattern is obtained by analyzing the noise source from the input signal and generating noise for a certain electric field strength. A pattern may be created.
• one of the causes of noise generation is noise that also generates circuit power used in the receiver.
• Noise with this circuit strength can be patterned according to the strength of the electric field because the circuit that causes the noise varies depending on the strength of the electric field. Therefore, in the present embodiment, the noise pattern for each electric field strength is preliminarily calculated and the noise pattern is calculated.
• the sound pattern is stored in the noise pattern storage unit 42, and when noise is suppressed, a necessary noise pattern is read from the noise pattern storage unit 42 to suppress noise.
• the electric field strength averaging unit 41 corresponds to the "signal strength average value calculation unit that averages the signal strength with the frame length formed by the frame forming means", and the time length (frame length for suppressing noise). ) To calculate the average electric field strength. Since the average value of the electric field strength is also calculated inside the electric field strength analysis unit 3, the electric field strength averaging unit 41 can be omitted.
• Fig. 10 is a graph showing the noise spectrum when the received electric field strength is OdB in the FM receiver with the FFT frame length changed. As shown in Fig. 10, the noise pattern stored in the noise pattern storage unit 42 is correlated to each noise spectrum when spectrum conversion is performed by changing the frame length of noise with the same electric field strength. It is known.
• the noise pattern storage unit 42 stores only the noise pattern for the longest frame length instead of storing the noise pattern for each frame length of frame formation.
• the noise calculation unit 43 creates a noise pattern by linear interpolation corresponding to the frame length or gain adjustment and thinning out the number of points.
• the characteristics with the highest spectral level are the characteristics at 2048-point FFT.
• the number of points is reduced to 1024 points, 512 points, and the one with the lowest spectral level is the 32-point FFT. It is a characteristic of time.
• the noise calculation unit 43 calculates the corresponding frame length. By generating the noise pattern, the memory capacity of the noise pattern storage unit 42 can be saved.
• FIG. 11 is a block diagram showing an example of a detailed configuration when the noise of the input signal is suppressed by subtracting the input spectrum force from the noise spectrum in the state of the frequency spectrum.
• the determined The converted frame is subjected to spectrum conversion by the spectrum conversion unit 2 (S307), and the noise estimation unit 4 estimates the noise spectrum of the converted frame (S308).
• the input signal spectrum output from the spectrum conversion unit 2 and the estimated noise vector output from the noise estimation unit 4 in this manner are input to the noise suppression unit 5, the amount of suppression is determined, and the noise included in the input signal is determined. The component is suppressed.
• the spectrum subtraction unit 52 subtracts the input signal spectrum power noise spectrum in the frequency domain (S309).
• the noise spectrum is also removed from the frequency vector force, and the signal is converted back to a time domain signal by the spectrum inverse converter 53 (S310), thereby obtaining an output signal with good noise quality with reduced noise.
• noise may be suppressed by subtracting the noise spectrum in the state of the frequency spectrum as described above, or the calculation is performed by inverse conversion to a signal on the time axis. It is good.
• FIG. 12 is a block diagram showing an example of a detailed configuration when noise suppression is performed after the noise spectrum and the input signal are inversely converted into a time domain signal.
• a convolution operation unit 54 is provided in the subsequent stage of the spectrum inverse conversion unit 53, and the convolution operation unit 54 performs a convolution operation with the input signal in the time domain.
• Patent Document 2 As in Patent Document 2,
• the suppression amount calculation unit 51 calculates the suppression amount as a suppression amount, it is more effective to control the suppression coefficient a and the threshold coefficient in consideration of the frame length calculated by the electric field strength analysis unit 3 including only the electric field strength. Noise suppression effect can be obtained. For example, even if the field strength is the same, if the frame length is short, it can be determined that the field strength is a dip in a multipath environment that is not caused by uniform degradation of the received field strength. There is a high possibility of non-stationary pulse noise, not white noise, and the amount of suppression can be increased by reducing the threshold coefficient. At this time, shorten the frame length.
• the estimated noise spectrum IN (f) I is an error even if it is estimated by a stationary noise pattern. It is possible to obtain an effective noise suppression effect that does not increase.
• the noise suppression device 100 of the present invention is useful for in-vehicle FM and AM radio receivers and television receivers.
• the present invention is not limited to this, but when incorporated in an on-vehicle receiver, the following special effects can be obtained. That is, in the case of an in-vehicle receiver, the electric field strength is likely to fluctuate frequently when the automobile travels, for example, in a valley of a building or a mountainous area. However, when an electric field strength of a certain level or higher is obtained, the influence of noise is not so great, so it is considered that the harmful effect is too much to follow the change in electric field strength.
• the threshold for determining the frame length is set to a value larger than that when the vehicle is stationary so that the vehicle does not follow the change in the electric field strength very sensitively.
• the frame length should be at least 256 points per frame when the car is running, and the maximum frame length is 1024 points when stationary.
• the electric field strength can be divided into one frame with a maximum of 1024 points of samples. it can.
• the frame length during travel and the frame length when stationary are not limited to the above examples (256 points during travel, 1024 points when stationary), but different numbers of points (for example, 128 points during travel and 2048 points when stationary) Etc.).
• the threshold when the received electric field strength is unrelated to the case of in-vehicle use, the threshold is set to a high value because it is not easily affected by noise, and when the received electric field strength is low, the threshold is set to a low value.
• the frame is fixed length to further improve the frequency resolution.
• the frame forming unit 1 of the present embodiment further states that “when the length of the formed frame is less than a predetermined length, a zero that is extended to a fixed length is added by adding a continuous zero to the formed frame. Equipped with the function of “expansion part”.
• FIG. 13 is a diagram illustrating an example of processing in the frame forming unit 1 when a fixed-length frame is formed by zero-filling both ends of the determined frame length. As shown in FIG.
• the frame forming unit 1 first determines the frame length specified by the field strength analysis unit 3 from the input signal. Form a frame. Next, it is further zero-extended to the length of the subsequent spectral conversion, 1024 points in the figure. In other words, the spectrum conversion is performed with a fixed length. In other words, each frame can be made to have a fixed length by performing zero extension, so that the spectrum converter 2 only needs to perform spectrum conversion of 1024 points at all times.
• the noise pattern storage unit 42 the input frequency spectrum power may be subtracted at the rate calculated by the noise calculation unit 43 which only needs to store the noise spectrum having a frame length of 1024 points. As a result, it is possible to perform noise suppression processing with very high computational efficiency. By zero extension, there is an effect that the frequency resolution at the time of spectrum conversion can be increased and the noise suppression performance can be further improved.
• FIG. 14 is a flowchart showing a procedure in the frame forming unit 1 when a lower limit is set for the frame length.
• the threshold is 256 in the figure.
• the frame forming unit 1 determines whether or not the frame length determined by the electric field strength analyzing unit 3 is 256 or more. Stretch the stretch frame.
• Spectral conversion unit 2 performs vector conversion with zero extension up to 256 points.
• the third embodiment differs greatly from the first embodiment in that the field strength is compared for each sample instead of comparing the field strength in units of averaged frames.
• the FIG. 15 is a block diagram showing an example of a configuration in the case where the field strength is compared for each sample without obtaining the average of the field strength in the averaged frame.
• the electric field intensity analyzing unit 3 does not include the electric field intensity averaging unit 31. That is, it is not necessary for the electric field intensity change detection unit 32 to compare the electric field intensity in units of averaged frames. For example, the electric field intensity may be compared for each sample.
• the frame length to be formed can be determined more dynamically, but the frame length may not be a power of 2 that is convenient for subsequent spectral conversion. Increases nature.
• the frame length determination unit 33 When correcting the frame length to a power of 2, the frame length determination unit 33
• It can be a configuration in which the sample at the boundary of the frame is moved back and forth so that the frame length is a power of 2.
• the electric field strength for determining the frame delimiter is divided into several zones. It may be divided into FIG. 16 is a diagram showing an example in which the electric field strength of the input signal is divided into several zones according to the noise pattern peculiar to the electric field strength.
• FIG. 17 is a flowchart showing an example in which the field strength of each sample of the input signal is examined and a new frame is formed each time the field strength zone changes. As shown in Fig. 16, the electric field strength that becomes the boundary of frame formation is preliminarily specified (Zone (1), Zone (2), Zone (3)), and the electric field strength force of a sample is determined. The frame length is determined when the field strength zone is straddled.
• the frame forming unit 1 checks the electric field strength (S1302), and checks the zone to which the electric field strength of the sample belongs (S1303).
• the first sample belongs to zone (3).
• the frame forming unit 1 checks whether or not the electric field intensity of the next sample has moved to a zone different from the electric field intensity of the immediately preceding sample (S1304).
• the field strength of the input signal is further reduced and is in the same zone (3) as the field strength of the previous sample. Therefore, the electric field strength analysis unit 3 Returns to the process of step S1301, and starts the process for the next sample.
• the field strength strength of the sample is examined and the comparison with the zone to which the immediately previous sample belongs is repeated (S1302, S1303).
• a predetermined number of samples or more belong to the same zone it may be divided into frames by a predetermined number of samples.
• the electric field strength of the input signal changes from a decrease to an increase.
• the electric field strength is examined in the next sample (S 1302), and the zone to which the electric field strength belongs is examined (S 1303).
• the electric field strength analysis unit 3 outputs the sample number n of the samples belonging to the same zone (3) until then to the frame forming unit 1, and the frame forming unit 1 forms the frame A.
• S 1305 One frame determined as a frame by the frame forming unit 1 is subjected to spectrum conversion by the subsequent spectrum conversion unit (S1306), and the noise spectrum of the pattern corresponding to the zone (zone (3)) is read from the noise estimation unit 4. (S1307).
• a process of suppressing the noise of the frame in zone (3) is performed (S 1308).
• the electric field strength analysis unit 3 returns to the process of step S1301, investigates the electric field strength and the zone to which each sample is input next (S1302, 1303), and the electric field strength of the input signal is the zone. It is checked whether or not it has moved from (2) to another zone (S 1304). If the electric field strength of the sample of the input signal has moved from zone (2) to another zone, for example, zone (1), the number of samples of samples belonging to zone (2) so far is framed. Notify Part 1 and form Frame B. Thereafter, the noise suppression device repeats the above processing until no input signal sample is input.
• the noise estimation procedure can be simplified and the amount of computation can be reduced. it can.
• FM radio it has been found that when the electric field strength is large, the low frequency spectrum increases as the force electric field strength, which is a white noise characteristic, decreases. Therefore, if a field strength zone is specified according to the noise characteristics of the low frequency range, noise estimation can be performed without a large error even with one noise pattern for each field strength zone.
• the application example of the noise suppression apparatus is described as a radio receiver, but it is useful for a television receiver or a mobile phone.
• this noise suppression device is a vehicle-mounted FM radio receiver that is highly likely to change the received electric field strength of radio waves with time due to geographical conditions while traveling and that requires good sound quality. It is more suitable for.
• any device that can know the electric field strength along with the input signal is useful as an application destination of this noise suppression device. That is, the present noise suppression apparatus can be implemented even using an index that represents the strength of the input signal instead of the electric field strength.
• FIG. 18 is a block diagram showing an example of a configuration of a noise suppression device that estimates noise patterns mixed in an input signal using a signal strength level, such as a voltage value, as a guide instead of electric field strength, and suppresses noise. is there. This figure assumes a playback device such as a CD or DVD.
• the noise suppression effect of the present invention can be obtained by using the signal strength s of the read signal measured by the signal strength analyzer 6. Can do.
• the noise suppression device of the present invention is a noise suppression device that can effectively suppress steady noise due to a decrease in electric field strength, and can also effectively suppress transient environmental noise. Useful.
• the receiver incorporating the noise suppression device of the present invention is useful as an FM radio receiver, an AM radio receiver, a television receiver, a mobile phone, and the like, particularly as an in-vehicle FM radio receiver. .

Landscapes

• Engineering & Computer Science (AREA)
• Signal Processing (AREA)
• Computational Linguistics (AREA)
• Quality & Reliability (AREA)
• Health & Medical Sciences (AREA)
• Audiology, Speech & Language Pathology (AREA)
• Human Computer Interaction (AREA)
• Physics & Mathematics (AREA)
• Acoustics & Sound (AREA)
• Multimedia (AREA)
• Computer Networks & Wireless Communication (AREA)
• Noise Elimination (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Publications (1)

Family

ID=36319064

Family Applications (1)

Country Status (6)

Cited By (4)

Families Citing this family (15)

Citations (2)

Family Cites Families (7)

• 2005
  • 2005-10-26 WO PCT/JP2005/019663 patent/WO2006049052A1/ja active Application Filing
  • 2005-10-26 DE DE602005026426T patent/DE602005026426D1/de active Active
  • 2005-10-26 US US11/665,641 patent/US7729680B2/en not_active Expired - Fee Related
  • 2005-10-26 CN CN2005800374437A patent/CN101053161B/zh not_active Expired - Fee Related
  • 2005-10-26 EP EP05799406A patent/EP1808965B1/en not_active Expired - Fee Related
  • 2005-10-26 JP JP2006543197A patent/JP4348369B2/ja not_active Expired - Fee Related

Patent Citations (2)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events